version 1.126, 2011/01/30 16:05:37 |
version 1.140, 2011/05/17 10:48:06 |
|
|
#include <unistd.h> |
#include <unistd.h> |
|
|
#include "mandoc.h" |
#include "mandoc.h" |
|
#include "libmandoc.h" |
#include "out.h" |
#include "out.h" |
#include "chars.h" |
|
#include "html.h" |
#include "html.h" |
#include "main.h" |
#include "main.h" |
|
|
Line 94 static const char *const htmlattrs[ATTR_MAX] = { |
|
Line 94 static const char *const htmlattrs[ATTR_MAX] = { |
|
"colspan", /* ATTR_COLSPAN */ |
"colspan", /* ATTR_COLSPAN */ |
}; |
}; |
|
|
|
static const char *const roffscales[SCALE_MAX] = { |
|
"cm", /* SCALE_CM */ |
|
"in", /* SCALE_IN */ |
|
"pc", /* SCALE_PC */ |
|
"pt", /* SCALE_PT */ |
|
"em", /* SCALE_EM */ |
|
"em", /* SCALE_MM */ |
|
"ex", /* SCALE_EN */ |
|
"ex", /* SCALE_BU */ |
|
"em", /* SCALE_VS */ |
|
"ex", /* SCALE_FS */ |
|
}; |
|
|
static void print_num(struct html *, const char *, size_t); |
static void print_num(struct html *, const char *, size_t); |
static void print_spec(struct html *, enum roffdeco, |
static void print_spec(struct html *, const char *, size_t); |
const char *, size_t); |
|
static void print_res(struct html *, const char *, size_t); |
static void print_res(struct html *, const char *, size_t); |
static void print_ctag(struct html *, enum htmltag); |
static void print_ctag(struct html *, enum htmltag); |
static void print_doctype(struct html *); |
static void print_doctype(struct html *); |
static void print_xmltype(struct html *); |
static void print_xmltype(struct html *); |
static int print_encode(struct html *, const char *, int); |
static int print_encode(struct html *, const char *, int); |
static void print_metaf(struct html *, enum roffdeco); |
static void print_metaf(struct html *, enum mandoc_esc); |
static void print_attr(struct html *, |
static void print_attr(struct html *, |
const char *, const char *); |
const char *, const char *); |
static void *ml_alloc(char *, enum htmltype); |
static void *ml_alloc(char *, enum htmltype); |
Line 120 ml_alloc(char *outopts, enum htmltype type) |
|
Line 132 ml_alloc(char *outopts, enum htmltype type) |
|
toks[2] = "includes"; |
toks[2] = "includes"; |
toks[3] = NULL; |
toks[3] = NULL; |
|
|
h = calloc(1, sizeof(struct html)); |
h = mandoc_calloc(1, sizeof(struct html)); |
if (NULL == h) { |
|
perror(NULL); |
|
exit((int)MANDOCLEVEL_SYSERR); |
|
} |
|
|
|
h->type = type; |
h->type = type; |
h->tags.head = NULL; |
h->tags.head = NULL; |
h->symtab = chars_init(CHARS_HTML); |
h->symtab = mchars_alloc(); |
|
|
while (outopts && *outopts) |
while (outopts && *outopts) |
switch (getsubopt(&outopts, UNCONST(toks), &v)) { |
switch (getsubopt(&outopts, UNCONST(toks), &v)) { |
Line 178 html_free(void *p) |
|
Line 186 html_free(void *p) |
|
} |
} |
|
|
if (h->symtab) |
if (h->symtab) |
chars_free(h->symtab); |
mchars_free(h->symtab); |
|
|
free(h); |
free(h); |
} |
} |
Line 214 print_gen_head(struct html *h) |
|
Line 222 print_gen_head(struct html *h) |
|
} |
} |
} |
} |
|
|
|
/* ARGSUSED */ |
static void |
static void |
print_num(struct html *h, const char *p, size_t len) |
print_num(struct html *h, const char *p, size_t len) |
{ |
{ |
const char *rhs; |
char c; |
|
|
rhs = chars_num2char(p, len); |
if ('\0' != (c = mchars_num2char(p, len))) |
if (rhs) |
putchar((int)c); |
putchar((int)*rhs); |
|
} |
} |
|
|
|
|
static void |
static void |
print_spec(struct html *h, enum roffdeco d, const char *p, size_t len) |
print_spec(struct html *h, const char *p, size_t len) |
{ |
{ |
int cp; |
int cp; |
const char *rhs; |
const char *rhs; |
size_t sz; |
size_t sz; |
|
|
if ((cp = chars_spec2cp(h->symtab, p, len)) > 0) { |
if ((cp = mchars_spec2cp(h->symtab, p, len)) > 0) { |
printf("&#%d;", cp); |
printf("&#%d;", cp); |
return; |
return; |
} else if (-1 == cp && DECO_SSPECIAL == d) { |
} else if (-1 == cp && 1 == len) { |
fwrite(p, 1, len, stdout); |
fwrite(p, 1, len, stdout); |
return; |
return; |
} else if (-1 == cp) |
} else if (-1 == cp) |
return; |
return; |
|
|
if (NULL != (rhs = chars_spec2str(h->symtab, p, len, &sz))) |
if (NULL != (rhs = mchars_spec2str(h->symtab, p, len, &sz))) |
fwrite(rhs, 1, sz, stdout); |
fwrite(rhs, 1, sz, stdout); |
} |
} |
|
|
Line 254 print_res(struct html *h, const char *p, size_t len) |
|
Line 260 print_res(struct html *h, const char *p, size_t len) |
|
const char *rhs; |
const char *rhs; |
size_t sz; |
size_t sz; |
|
|
if ((cp = chars_res2cp(h->symtab, p, len)) > 0) { |
if ((cp = mchars_res2cp(h->symtab, p, len)) > 0) { |
printf("&#%d;", cp); |
printf("&#%d;", cp); |
return; |
return; |
} else if (-1 == cp) |
} else if (-1 == cp) |
return; |
return; |
|
|
if (NULL != (rhs = chars_res2str(h->symtab, p, len, &sz))) |
if (NULL != (rhs = mchars_res2str(h->symtab, p, len, &sz))) |
fwrite(rhs, 1, sz, stdout); |
fwrite(rhs, 1, sz, stdout); |
} |
} |
|
|
|
|
static void |
static void |
print_metaf(struct html *h, enum roffdeco deco) |
print_metaf(struct html *h, enum mandoc_esc deco) |
{ |
{ |
enum htmlfont font; |
enum htmlfont font; |
|
|
switch (deco) { |
switch (deco) { |
case (DECO_PREVIOUS): |
case (ESCAPE_FONTPREV): |
font = h->metal; |
font = h->metal; |
break; |
break; |
case (DECO_ITALIC): |
case (ESCAPE_FONTITALIC): |
font = HTMLFONT_ITALIC; |
font = HTMLFONT_ITALIC; |
break; |
break; |
case (DECO_BOLD): |
case (ESCAPE_FONTBOLD): |
font = HTMLFONT_BOLD; |
font = HTMLFONT_BOLD; |
break; |
break; |
case (DECO_ROMAN): |
case (ESCAPE_FONTROMAN): |
font = HTMLFONT_NONE; |
font = HTMLFONT_NONE; |
break; |
break; |
default: |
default: |
Line 302 print_metaf(struct html *h, enum roffdeco deco) |
|
Line 308 print_metaf(struct html *h, enum roffdeco deco) |
|
print_otag(h, TAG_I, 0, NULL); |
print_otag(h, TAG_I, 0, NULL); |
} |
} |
|
|
|
int |
|
html_strlen(const char *cp) |
|
{ |
|
int ssz, sz; |
|
const char *seq, *p; |
|
|
|
/* |
|
* Account for escaped sequences within string length |
|
* calculations. This follows the logic in term_strlen() as we |
|
* must calculate the width of produced strings. |
|
* Assume that characters are always width of "1". This is |
|
* hacky, but it gets the job done for approximation of widths. |
|
*/ |
|
|
|
sz = 0; |
|
while (NULL != (p = strchr(cp, '\\'))) { |
|
sz += (int)(p - cp); |
|
++cp; |
|
switch (mandoc_escape(&cp, &seq, &ssz)) { |
|
case (ESCAPE_ERROR): |
|
return(sz); |
|
case (ESCAPE_NUMBERED): |
|
/* FALLTHROUGH */ |
|
case (ESCAPE_PREDEF): |
|
/* FALLTHROUGH */ |
|
case (ESCAPE_SPECIAL): |
|
sz++; |
|
break; |
|
default: |
|
break; |
|
} |
|
} |
|
|
|
assert(sz >= 0); |
|
return(sz + strlen(cp)); |
|
} |
|
|
static int |
static int |
print_encode(struct html *h, const char *p, int norecurse) |
print_encode(struct html *h, const char *p, int norecurse) |
{ |
{ |
size_t sz; |
size_t sz; |
int len, nospace; |
int len, nospace; |
const char *seq; |
const char *seq; |
enum roffdeco deco; |
enum mandoc_esc esc; |
static const char rejs[6] = { '\\', '<', '>', '&', ASCII_HYPH, '\0' }; |
static const char rejs[6] = { '\\', '<', '>', '&', ASCII_HYPH, '\0' }; |
|
|
nospace = 0; |
nospace = 0; |
|
|
for (; *p; p++) { |
while ('\0' != *p) { |
sz = strcspn(p, rejs); |
sz = strcspn(p, rejs); |
|
|
fwrite(p, 1, sz, stdout); |
fwrite(p, 1, sz, stdout); |
p += /* LINTED */ |
p += (int)sz; |
sz; |
|
|
|
if ('<' == *p) { |
if ('\0' == *p) |
|
break; |
|
|
|
switch (*p++) { |
|
case ('<'): |
printf("<"); |
printf("<"); |
continue; |
continue; |
} else if ('>' == *p) { |
case ('>'): |
printf(">"); |
printf(">"); |
continue; |
continue; |
} else if ('&' == *p) { |
case ('&'): |
printf("&"); |
printf("&"); |
continue; |
continue; |
} else if (ASCII_HYPH == *p) { |
case (ASCII_HYPH): |
/* |
|
* Note: "soft hyphens" aren't graphically |
|
* displayed when not breaking the text; we want |
|
* them to be displayed. |
|
*/ |
|
/*printf("­");*/ |
|
putchar('-'); |
putchar('-'); |
continue; |
continue; |
} else if ('\0' == *p) |
default: |
break; |
break; |
|
} |
|
|
seq = ++p; |
esc = mandoc_escape(&p, &seq, &len); |
len = a2roffdeco(&deco, &seq, &sz); |
if (ESCAPE_ERROR == esc) |
|
break; |
|
|
switch (deco) { |
switch (esc) { |
case (DECO_NUMBERED): |
case (ESCAPE_NUMBERED): |
print_num(h, seq, sz); |
print_num(h, seq, len); |
break; |
break; |
case (DECO_RESERVED): |
case (ESCAPE_PREDEF): |
print_res(h, seq, sz); |
print_res(h, seq, len); |
break; |
break; |
case (DECO_SSPECIAL): |
case (ESCAPE_SPECIAL): |
/* FALLTHROUGH */ |
print_spec(h, seq, len); |
case (DECO_SPECIAL): |
|
print_spec(h, deco, seq, sz); |
|
break; |
break; |
case (DECO_PREVIOUS): |
case (ESCAPE_FONTPREV): |
/* FALLTHROUGH */ |
/* FALLTHROUGH */ |
case (DECO_BOLD): |
case (ESCAPE_FONTBOLD): |
/* FALLTHROUGH */ |
/* FALLTHROUGH */ |
case (DECO_ITALIC): |
case (ESCAPE_FONTITALIC): |
/* FALLTHROUGH */ |
/* FALLTHROUGH */ |
case (DECO_ROMAN): |
case (ESCAPE_FONTROMAN): |
if (norecurse) |
if (norecurse) |
break; |
break; |
print_metaf(h, deco); |
print_metaf(h, esc); |
break; |
break; |
|
case (ESCAPE_NOSPACE): |
|
if ('\0' == *p) |
|
nospace = 1; |
|
break; |
default: |
default: |
break; |
break; |
} |
} |
|
|
p += len - 1; |
|
|
|
if (DECO_NOSPACE == deco && '\0' == *(p + 1)) |
|
nospace = 1; |
|
} |
} |
|
|
return(nospace); |
return(nospace); |
Line 401 print_otag(struct html *h, enum htmltag tag, |
|
Line 439 print_otag(struct html *h, enum htmltag tag, |
|
/* Push this tags onto the stack of open scopes. */ |
/* Push this tags onto the stack of open scopes. */ |
|
|
if ( ! (HTML_NOSTACK & htmltags[tag].flags)) { |
if ( ! (HTML_NOSTACK & htmltags[tag].flags)) { |
t = malloc(sizeof(struct tag)); |
t = mandoc_malloc(sizeof(struct tag)); |
if (NULL == t) { |
|
perror(NULL); |
|
exit((int)MANDOCLEVEL_SYSERR); |
|
} |
|
t->tag = tag; |
t->tag = tag; |
t->next = h->tags.head; |
t->next = h->tags.head; |
h->tags.head = t; |
h->tags.head = t; |
Line 442 print_otag(struct html *h, enum htmltag tag, |
|
Line 476 print_otag(struct html *h, enum htmltag tag, |
|
print_attr(h, "lang", "en"); |
print_attr(h, "lang", "en"); |
} |
} |
|
|
/* Accomodate for XML "well-formed" singleton escaping. */ |
/* Accommodate for XML "well-formed" singleton escaping. */ |
|
|
if (HTML_AUTOCLOSE & htmltags[tag].flags) |
if (HTML_AUTOCLOSE & htmltags[tag].flags) |
switch (h->type) { |
switch (h->type) { |
Line 518 print_doctype(struct html *h) |
|
Line 552 print_doctype(struct html *h) |
|
name, doctype, dtd); |
name, doctype, dtd); |
} |
} |
|
|
|
|
void |
void |
print_text(struct html *h, const char *word) |
print_text(struct html *h, const char *word) |
{ |
{ |
|
|
if (word[0] && '\0' == word[1]) |
|
switch (word[0]) { |
|
case('.'): |
|
/* FALLTHROUGH */ |
|
case(','): |
|
/* FALLTHROUGH */ |
|
case(';'): |
|
/* FALLTHROUGH */ |
|
case(':'): |
|
/* FALLTHROUGH */ |
|
case('?'): |
|
/* FALLTHROUGH */ |
|
case('!'): |
|
/* FALLTHROUGH */ |
|
case(')'): |
|
/* FALLTHROUGH */ |
|
case(']'): |
|
if ( ! (HTML_IGNDELIM & h->flags)) |
|
h->flags |= HTML_NOSPACE; |
|
break; |
|
default: |
|
break; |
|
} |
|
|
|
if ( ! (HTML_NOSPACE & h->flags)) { |
if ( ! (HTML_NOSPACE & h->flags)) { |
/* Manage keeps! */ |
/* Manage keeps! */ |
if ( ! (HTML_KEEP & h->flags)) { |
if ( ! (HTML_KEEP & h->flags)) { |
Line 574 print_text(struct html *h, const char *word) |
|
Line 583 print_text(struct html *h, const char *word) |
|
} |
} |
|
|
h->flags &= ~HTML_IGNDELIM; |
h->flags &= ~HTML_IGNDELIM; |
|
|
/* |
|
* Note that we don't process the pipe: the parser sees it as |
|
* punctuation, but we don't in terms of typography. |
|
*/ |
|
if (word[0] && '\0' == word[1]) |
|
switch (word[0]) { |
|
case('('): |
|
/* FALLTHROUGH */ |
|
case('['): |
|
h->flags |= HTML_NOSPACE; |
|
break; |
|
default: |
|
break; |
|
} |
|
} |
} |
|
|
|
|
|
|
bufcat_su(struct html *h, const char *p, const struct roffsu *su) |
bufcat_su(struct html *h, const char *p, const struct roffsu *su) |
{ |
{ |
double v; |
double v; |
const char *u; |
|
|
|
v = su->scale; |
v = su->scale; |
|
if (SCALE_MM == su->unit && 0.0 == (v /= 100.0)) |
|
v = 1.0; |
|
|
switch (su->unit) { |
buffmt(h, "%s: %.2f%s;", p, v, roffscales[su->unit]); |
case (SCALE_CM): |
|
u = "cm"; |
|
break; |
|
case (SCALE_IN): |
|
u = "in"; |
|
break; |
|
case (SCALE_PC): |
|
u = "pc"; |
|
break; |
|
case (SCALE_PT): |
|
u = "pt"; |
|
break; |
|
case (SCALE_EM): |
|
u = "em"; |
|
break; |
|
case (SCALE_MM): |
|
if (0 == (v /= 100)) |
|
v = 1; |
|
u = "em"; |
|
break; |
|
case (SCALE_EN): |
|
u = "ex"; |
|
break; |
|
case (SCALE_BU): |
|
u = "ex"; |
|
break; |
|
case (SCALE_VS): |
|
u = "em"; |
|
break; |
|
default: |
|
u = "ex"; |
|
break; |
|
} |
|
|
|
/* |
|
* XXX: the CSS spec isn't clear as to which types accept |
|
* integer or real numbers, so we just make them all decimals. |
|
*/ |
|
buffmt(h, "%s: %.2f%s;", p, v, u); |
|
} |
} |
|
|
|
|